The multiple function printer (MFP) and scanner play essential role to produce hot copy image and digital image. To provide image with high fidelity, an image processing software is required to compensate the image obtained through the MFP and scanner.
FIG. 1 shows a prior art image processing method. The original document is scanned to obtain digitalized RGB data. The digitalized RGB data is subjected to a CMYK conversion, a half-tone treatment and a graphic-text enhancing treatment to obtain CMYK half-tone image. The CMYK half-tone image is printed by a printer module. As can be seen from the flowchart, the scanned image is processed by the CMYK conversion, the half-tone treatment and the graphic-text enhancing treatment to compensate the discrepancy between scanner and printer.
The half-tone treatment generally adopts order dither and error diffusion methods. The order dither increases gray level by decreasing the resolution of image, i.e., increases the coding point in a region for a specific color to increase the gray level for this color.
The error diffusion method distributes the error during digitalizing the gray scale image to neighbor pixel. With reference to FIG. 2, the error diffusion method is implemented by a first adder 11, a quantizer circuit 12, a second adder 13 and an error filter 14, wherein Xij is input image pixel, Uij is error diffusion pixel, X′ij is output image pixel, eij is neighbor image error, H(z) is error filter. The second adder 13 subtracts the error diffusion pixel Uij from the output image pixel X′ij to obtain the neighbor image error eij. The neighbor image error eij is processed by the error filter H(z) 14 to obtain a corrected pixel error H(e(i,j)). The first adder 11 subtracts the corrected pixel error H(e(i,j)) from the input image pixel Xij to obtain the error diffusion pixel Uij. It also means that the pixel error H(e(i,j)) is distributed to the neighbor pixels according to coefficients type of the error filter H(z). The error diffusion pixel Uij is compared with a threshold t in the quantizer circuit 12 to obtain digitalized output (0 or 1). A digital output 1 is obtained when the error diffusion pixel Uij is larger than the threshold t and a digital output 0 is obtained when the error diffusion pixel Uij is smaller than or equal to the threshold t. The error diffusion method distributes quantization error to neighbor pixel to compensate error between one pixel and neighbor pixel.
The conventional half-tone treatment such as order dither and error diffusion methods generally produces moire in scanned image. Therefore, post treatment such as mask method or frequency domain method such as Fourier transform is used to reduce the moire in scanned image. The about methods become time consuming and complicated when the moire frequently occur or occupies large area.